Dynamic Changes Of Vegetation Coverage And Its Response To Climate Change Of Loess Plateau In Northern Shaanxi Province Based On ESTARFM Mode

The Loess Plateau in Northern Shaanxi Province has a complex topography and a fragile ecological environment,and it is a serious area of soil and water loss in China.Fractional vegetation cover(FVC)is an important factor for evaluating regional ecological environment,so the study with the spatio-temporal dynamic changes of the vegetation coverage has a great significance for the effective measures of soil and water conservation,ecological environmental engineering in this region.With the continuous development of space satellite science,remote sensing data has become the main data for estimating regional vegetation coverage.However,the existing remote sensing data acquired by a sensor is restricted by the weather and satellite return period or other factors,thus as a data source,the FVC estimation has a limitiation to meet the requirements of high spatial and temporal resolution data.Using spatial-temporal fusion technology produces the high spatial and temporal resolution data that provides a new idea and method for study of vegetation cover in the complex terrain.We used Landsat and MODIS data as data sources,based on ESTARFM mode to fuse the Landsat NDVI data in the Loess Plateau in Northern Shaanxi Province from 2008 to2016(between June to August),and calculate FVC to analysis the spatial and temporal variation characteristics of vegetation coverage in the study area during 9 years.Then we analyzed the vegetation cover change that response to climate change using the meteorological data from 2008 to 2016,so as to provide scientific references for soil and water conservation and environmental sustainable development in Loess Plateau in Northern Shaanxi Province.The main research results are as follows:(1)High temporal and spatial FVC datas' s generation.Using Landsat and MODIS image data as data sources,then ues the ESTARFM spatial-temporal fusion model to make two scale NDVI vegetation index data fusion,get high spatial and tempora ESTARFM NDVI data(30m,8day).Then we use actual Landsat NDVI data with fusion results to analyse the effectiveness.The results showed that ESTARFM NDVI was highly correlated with Landsat NDVI in the same period after visual interpretation.After quantitative analysis,the overallcorrelation coefficient is above 0.97,the difference of most pixels value was close to 0,the overall fusion accuracy is higher.Finally,the obtained Landsat NDVI time series data from June to August,2008 and 2016 were used to estimate vegetation coverage for subsequent studies.(2)Spatial and temporal variation analysis of vegetation cover in Loess Plateau in Northern Shaanxi Province that based on high spatial and temporal vegetation data.In terms of inter-annual changes: the vegetation coverage of the whole region showed an increasing trend from 2008 to 2016(between June to August).The area proportion of vegetation evolution is much larger than that of degeneration,and the vegetation area proportion of high grade vegetation cover type is increasing.The vegetation change of each land use types are rising,among which the shrubbery and the open woodland are the most obvious rise.In terms of space,the spatial distribution of vegetation coverage in the whole region is characterized by a high level in the southeast and a low level in the northwest,and it is improving year by year.The regions that enjoy obvious improvement effects are mainly distributed in counties or districts north of the research areas of Yanchang and Ansai and south of Fugu and Shenmu.The overall restoration effect of vegetation are good.(3)Vegetation coverage changes response to climate factors in the Loess Plateau in Northern Shaanxi Province.From the perspective of spatial correlation,the spatial difference characteristics of vegetation coverage response to climatic factors in the same season and the preceding season in June and August were discussed.The influence of temperature and solar radiation on vegetation is mainly distributed in the central and northern part of the study area.The influence of wind speed on vegetation was mainly distributed in the central and southern regions,and the influence range of the same season was larger than that of the previous season.The result of the overall analysis of vegetation's response to the climate and its time-lag from June to August from the perspective of time correlations shows that vegetation has a lag phase of 1 to 3 months in response to air temperature,rainfall and relative humidity,but its response to wind speed and solar radiation is not salient.